Accuracy and generalization capability of an automatic method for the detection of typical brain hypometabolism in prodromal Alzheimer disease

for the Alzheimer’s Disease Neuroimaging Initiative

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Purpose: The aim of this study was to verify the reliability and generalizability of an automatic tool for the detection of Alzheimer-related hypometabolic pattern based on a Support-Vector-Machine (SVM) model analyzing 18F-fluorodeoxyglucose (FDG) PET data. Methods: The SVM model processed metabolic data from anatomical volumes of interest also considering interhemispheric asymmetries. It was trained on a homogeneous dataset from a memory clinic center and tested on an independent multicentric dataset drawn from the Alzheimer’s Disease Neuroimaging Initiative. Subjects were included in the study and classified based on a diagnosis confirmed after an adequate follow-up time. Results: The accuracy of the discrimination between patients with Alzheimer Disease (AD), in either prodromal or dementia stage, and normal aging subjects was 95.8%, after cross-validation, in the training set. The accuracy of the same model in the testing set was 86.5%. The role of the two datasets was then reversed, and the accuracy was 89.8% in the multicentric training set and 88.0% in the monocentric testing set. The classification rate was also evaluated in different subgroups, including non-converter mild cognitive impairment (MCI) patients, subjects with MCI reverted to normal conditions and subjects with non-confirmed memory concern. The percent of pattern detections increased from 77% in early prodromal AD to 91% in AD dementia, while it was about 10% for healthy controls and non-AD patients. Conclusions: The present findings show a good level of reproducibility and generalizability of a model for detecting the hypometabolic pattern in AD and confirm the accuracy of FDG-PET in Alzheimer disease.

Original languageEnglish
JournalEuropean Journal of Nuclear Medicine and Molecular Imaging
DOIs
Publication statusAccepted/In press - Jan 1 2018

Fingerprint

Alzheimer Disease
Brain
Fluorodeoxyglucose F18
Generalization (Psychology)
Neuroimaging
Dementia
Datasets
Cognitive Dysfunction
Support Vector Machine

Keywords

  • Alzheimer disease
  • Classification and prediction
  • Discriminant analysis
  • FDG-PET
  • MCI due to AD
  • Neurodegenerative disorders
  • Neuroimage classification
  • Support vector machine

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

@article{0864b52be98a41a8835ad4b44ec96f9e,
title = "Accuracy and generalization capability of an automatic method for the detection of typical brain hypometabolism in prodromal Alzheimer disease",
abstract = "Purpose: The aim of this study was to verify the reliability and generalizability of an automatic tool for the detection of Alzheimer-related hypometabolic pattern based on a Support-Vector-Machine (SVM) model analyzing 18F-fluorodeoxyglucose (FDG) PET data. Methods: The SVM model processed metabolic data from anatomical volumes of interest also considering interhemispheric asymmetries. It was trained on a homogeneous dataset from a memory clinic center and tested on an independent multicentric dataset drawn from the Alzheimer’s Disease Neuroimaging Initiative. Subjects were included in the study and classified based on a diagnosis confirmed after an adequate follow-up time. Results: The accuracy of the discrimination between patients with Alzheimer Disease (AD), in either prodromal or dementia stage, and normal aging subjects was 95.8{\%}, after cross-validation, in the training set. The accuracy of the same model in the testing set was 86.5{\%}. The role of the two datasets was then reversed, and the accuracy was 89.8{\%} in the multicentric training set and 88.0{\%} in the monocentric testing set. The classification rate was also evaluated in different subgroups, including non-converter mild cognitive impairment (MCI) patients, subjects with MCI reverted to normal conditions and subjects with non-confirmed memory concern. The percent of pattern detections increased from 77{\%} in early prodromal AD to 91{\%} in AD dementia, while it was about 10{\%} for healthy controls and non-AD patients. Conclusions: The present findings show a good level of reproducibility and generalizability of a model for detecting the hypometabolic pattern in AD and confirm the accuracy of FDG-PET in Alzheimer disease.",
keywords = "Alzheimer disease, Classification and prediction, Discriminant analysis, FDG-PET, MCI due to AD, Neurodegenerative disorders, Neuroimage classification, Support vector machine",
author = "{for the Alzheimer’s Disease Neuroimaging Initiative} and {De Carli}, Fabrizio and Flavio Nobili and Marco Pagani and Matteo Bauckneht and Federico Massa and Matteo Grazzini and Cathrine Jonsson and Enrico Peira and Silvia Morbelli and Dario Arnaldi",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/s00259-018-4197-7",
language = "English",
journal = "European Journal of Pediatrics",
issn = "0340-6199",
publisher = "Springer Berlin Heidelberg",

}

TY - JOUR

T1 - Accuracy and generalization capability of an automatic method for the detection of typical brain hypometabolism in prodromal Alzheimer disease

AU - for the Alzheimer’s Disease Neuroimaging Initiative

AU - De Carli, Fabrizio

AU - Nobili, Flavio

AU - Pagani, Marco

AU - Bauckneht, Matteo

AU - Massa, Federico

AU - Grazzini, Matteo

AU - Jonsson, Cathrine

AU - Peira, Enrico

AU - Morbelli, Silvia

AU - Arnaldi, Dario

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Purpose: The aim of this study was to verify the reliability and generalizability of an automatic tool for the detection of Alzheimer-related hypometabolic pattern based on a Support-Vector-Machine (SVM) model analyzing 18F-fluorodeoxyglucose (FDG) PET data. Methods: The SVM model processed metabolic data from anatomical volumes of interest also considering interhemispheric asymmetries. It was trained on a homogeneous dataset from a memory clinic center and tested on an independent multicentric dataset drawn from the Alzheimer’s Disease Neuroimaging Initiative. Subjects were included in the study and classified based on a diagnosis confirmed after an adequate follow-up time. Results: The accuracy of the discrimination between patients with Alzheimer Disease (AD), in either prodromal or dementia stage, and normal aging subjects was 95.8%, after cross-validation, in the training set. The accuracy of the same model in the testing set was 86.5%. The role of the two datasets was then reversed, and the accuracy was 89.8% in the multicentric training set and 88.0% in the monocentric testing set. The classification rate was also evaluated in different subgroups, including non-converter mild cognitive impairment (MCI) patients, subjects with MCI reverted to normal conditions and subjects with non-confirmed memory concern. The percent of pattern detections increased from 77% in early prodromal AD to 91% in AD dementia, while it was about 10% for healthy controls and non-AD patients. Conclusions: The present findings show a good level of reproducibility and generalizability of a model for detecting the hypometabolic pattern in AD and confirm the accuracy of FDG-PET in Alzheimer disease.

AB - Purpose: The aim of this study was to verify the reliability and generalizability of an automatic tool for the detection of Alzheimer-related hypometabolic pattern based on a Support-Vector-Machine (SVM) model analyzing 18F-fluorodeoxyglucose (FDG) PET data. Methods: The SVM model processed metabolic data from anatomical volumes of interest also considering interhemispheric asymmetries. It was trained on a homogeneous dataset from a memory clinic center and tested on an independent multicentric dataset drawn from the Alzheimer’s Disease Neuroimaging Initiative. Subjects were included in the study and classified based on a diagnosis confirmed after an adequate follow-up time. Results: The accuracy of the discrimination between patients with Alzheimer Disease (AD), in either prodromal or dementia stage, and normal aging subjects was 95.8%, after cross-validation, in the training set. The accuracy of the same model in the testing set was 86.5%. The role of the two datasets was then reversed, and the accuracy was 89.8% in the multicentric training set and 88.0% in the monocentric testing set. The classification rate was also evaluated in different subgroups, including non-converter mild cognitive impairment (MCI) patients, subjects with MCI reverted to normal conditions and subjects with non-confirmed memory concern. The percent of pattern detections increased from 77% in early prodromal AD to 91% in AD dementia, while it was about 10% for healthy controls and non-AD patients. Conclusions: The present findings show a good level of reproducibility and generalizability of a model for detecting the hypometabolic pattern in AD and confirm the accuracy of FDG-PET in Alzheimer disease.

KW - Alzheimer disease

KW - Classification and prediction

KW - Discriminant analysis

KW - FDG-PET

KW - MCI due to AD

KW - Neurodegenerative disorders

KW - Neuroimage classification

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=85055986240&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055986240&partnerID=8YFLogxK

U2 - 10.1007/s00259-018-4197-7

DO - 10.1007/s00259-018-4197-7

M3 - Article

AN - SCOPUS:85055986240

JO - European Journal of Pediatrics

JF - European Journal of Pediatrics

SN - 0340-6199

ER -